On Apr 6, 10:03 pm, Canny <[email protected]> wrote:
> On Apr 6, 9:11 pm, John J Barton <[email protected]> wrote:
>
> > On Apr 6, 5:51 pm, Canny <[email protected]> wrote:
>
> > > I have some ideas about creating new firebug debugging features. For
> > > example, given a downloaded web page, beautify (www.jsbeautifier.org)
> > > various contents (html/javascript/css) before the user is allowed to
> > > debug the web application. This should be very useful to work with
> > > very "ugly"-formatted third party contents. As another example, I want
>
> > To be successful you have to intercept the js code before firefox
> > compiles it. You can get most cases by intercepting net traffic (see
> > net.js and component/firebug-channel-listener.js or firebug-http-
> > observer.js)  but recognizing JS requires parsing HTML at least to
> > some simple degree.
>
> Since the Script panel does display JavaScript code, is there a string
> object representing the displayed JS code? Can I just pass that
> string object to jsbeautifier to obtain a well-formatted JS code
> string to display, given that the transformation happens before any
> other
> processing on the string such as syntax coloring?

Sure, but the line numbers will be all wrong and Firebug will not work
correctly as a result.

>
> > > to develop an automated tool that produces traces (an example trace
> > > could be the intermediate presentation of all of the executed
> > > javascript bytecode instructions along with their parameter values)
>
> > You can trace the function calls easily but producing a useful UI for
> > a complex application is difficult. We have no access to the
> > bytecodes.
>
> I don't quite need a complex UI. Just want to allow firebug to record
> the trace
> without human intervention. It would also be fine even if the JS
> bytecode level
> detail is not available in the trace. JavaScript source code level
> trace (e.g., recording
> every JS statement that has been executed, along with the values
> involved in
> executing the statement) can also be very useful. Even a trace of all
> functions
> being invoked could be very useful, but it's better be the case that
> functions in the trace is
> identified by some unique internal IDs, rather than superficial alias
> names. In addition, it's
> better be the case that the values of functions' true parameters can
> also be recorded in the
> trace, so that it's easier to see how certain values are used by the
> program.

Well I don't understand how this can be interesting without human
intervention. What is the purpose of a trace no one reads?

>
> -- Yan

-- 
You received this message because you are subscribed to the Google Groups 
"Firebug" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/firebug?hl=en.

Reply via email to